The long range goal of our research is to determine the most effective and efficient treatment interventions for alcohol abusers. Our working assumption is that a matching hypothesis is necessary: patients with different characteristics will respond differentially to different treatment interventions. Our approach to the problem has been to develop a theoretical formulation for making prognoses concerning outcome: Five variables are identified. These are alcohol dependence, psychological health, social dependence, social environmental support for alcohol health and social environmental support for psychological health. The theory includes hypotheses concerning the set of conditions under which specific treatment interventions will be differentially effective. Three outpatient treatment interventions are defined: (1) individual focus, a social learning theory approach to helping alcohol abusers to learn how to achieve and maintain abstinence; (2) enhancement of social support, an intervention which includes incorporation of significant others into the patients treatment program; and (3) enhancement of vocational role, an intervention which is added to individual focus and enhancement of social support. To test the theory and the differential effectiveness of the interventions, patients discharged from inpatient service are randomly assigned to three treatment conditions. They are treated for 20 outpatient sessions over the course of a year and are followed as research subjects for 18 months after treatment initiation. Interviews and comprehensive measurements are made of their alcohol, physical, psychological, social and vocational health. Treatment and changes in prognostic variables are also measured, in order to test the goodness of fit of the theoretical formulation.